Bengaluru-based Sarvam AI has unveiled “Chanakya,” a new artificial intelligence platform built specifically for high-security and mission-critical environments. The system is designed for sectors where data privacy, operational control, and reliability are essential, including defence, government, healthcare, and critical infrastructure.
Unlike conventional AI solutions that rely on cloud-based infrastructure, Chanakya is engineered to operate on-premise or within air-gapped systems, allowing it to function without any internet connectivity. This approach ensures that sensitive data remains within controlled environments, addressing growing concerns around data exposure and cybersecurity risks.
The platform is capable of handling complex, real-world applications by processing multiple forms of data, including text, images, and structured datasets. It also supports advanced workflows such as AI agents that can execute multi-step operations, positioning it beyond traditional chatbot or language model capabilities and into a full-stack AI system for decision-making and operational support.
Founded in 2023 by Vivek Raghavan, Pratyush Kumar, and Ankush Sabharwal, Sarvam AI has rapidly gained recognition in India’s artificial intelligence ecosystem. The company initially drew attention for developing large language models tailored to Indian languages and datasets, including Sarvam-30B and Sarvam-105B, as part of its effort to build culturally and linguistically relevant AI solutions.
With Chanakya, the company is expanding its focus toward what it describes as “high-trust” environments. The platform reflects a strategic shift to address enterprise and government demand for secure, accountable, and high-performance AI systems capable of operating in sensitive and regulated settings.
